【摘要】 实际图像采集处理系统中,经常遇到需要将多幅具有重叠区域的彩色图像自动拼接成一幅大的图像的难题。基于模板匹配的原理,文中提出了一种新的彩色图像自动拼接方法。对两幅具有重叠区域的彩色图像,该方法先利用图像特征信息从一幅图像的重叠区域中自动寻找一小块模板图像,然后根据最大相似性准则在另一幅图像中重叠区域里搜索,直至找到最佳图像配准点,最后利用平滑因子对两幅图像的重叠区域进行数据融合操作,实现了彩色图像的快速自动拼接。实践证明,该方法对大部分彩色图像具有很好的拼接效果。更多还原

【Abstract】 There is a very difficult problem on mosaicing multiple overlapped color images into one automatically in a practical image acquiring and processing system. A new algorithm based on Template-Matching is proposed in this paper. To mosaic two overlapped color images fast and automatically, our algorithm firstly acquire a small template image automatically from one image according to image attribute, then search a best matching point in another image, finally merge overlapped image area data by using smooth factors. A lot of experiments show that our algorithm is effective for most overlapped color images.更多还原